29.03.2024

BitBank.nz Provides Live Cryptocurency Forecasts with Machine Learning

The company was founded when Lee Penkman in between his day job as Motion RnD Technical Director at WetaFX and a family expecting baby #2 inspired by the Andrew Ng machine learning Coursera course, created a high-frequency cryptocurrency trading bot.

Needing a way of visualising the system running he created a live stats dashboard and saw a gap in the market between a fully automated trading bot and live charts with indicators to help traders make informed decisions.

So was born a forecasting system that enables you to peer into the mind of an automated trading bot.

Seeing the product in action Lee now believes the charts that exchanges show are “so light on details they encourage people to panic trade” and “Never show enough detail about whats currently happening and where the markets going, especially in crypto where we see extreme volatility and sudden swings”

At Weta Digital Lee was credited in Valerian and the City of a Thousand Planets a big budget visual effects movie.

Husband Wife, Lee Nicole and daughter with Weta Digitals Best Visual Effects Award For Jungle Book

Currently BitBank.nz shows forecasts up to 2 hours into the future for multiple cryptocurrencies across 70+ markets, it shows what prices the algorithm recommends buying and selling at and what projected profits there are after the exchange fees.

BitBank.nz has an API so that people can get live updates from the forecasting feed to make automated decisions/create custom charts and you can get bulk data for backtesting the accuracy of the system or finding correlations between BitBank.nz’s features.

BitBank.nz shows what its recommended buy/sell prices are given the projections and how much it expects to earn on them after .25% trading fees.

Lee says

“We are transparent about our forecasting accuracy so people can see our historical predictions and how good/bad they where simply by clicking on the graphs in the past”

BitBank.nz is just getting off the ground as we speak “We have paying customers, a good conversion rate which proves to me we have a strong product but we havn’t had the traffic or media attention that other enviable crypto companies or ICO’s we have seen recently”

Lee said “its difficult to market a plain SaaS company in the crypto space when your competing against ICO’s that promise parties on the moon and 100x returns, our main customers are pro traders, they realise that our price point is a fairly affordable middle-ground between getting someone else to manage your money and manually guessing the markets yourself, if you manage your own investments with some insight behind you that can go a long way”

For someone who operates a manual trading forecasts company, Lee is still optimistic about a fully automated trading approach “There is money to be made trading any level right now, with going fully automated like our competitors (cryptotrader) its hard to build a users trust in an automated system, i found when building BitBank.nz as an automated system you would quickly loose trust based on a single bad trade even if the forecasts where okay. Even if it was making money you’d look at the coin that doubled and retroactively say well why didn’t it buy that instead?”

Before founding https://BitBank.nz Lee got into Artificial Intelligence and Machine Learning by creating computer opponents for web based games like bigmultiplayerches.com and wordsmashing.com which managed to land him a job in Visual Effects.

“I had many learning successes, aka commercial failures before getting into crypto and eventually starting BitBank.nz I knew Artificial Intelligence was something that would be great to know in the coming years much like crypto, i didn’t know what i would end up doing with the skills as i learned about crypto and AI, i’m glad i found a place where i can combine the two. Its allways great to ride the wave of hype and new technology”

BitBank.nz predicts the weighted trade average in the next 5 minutes, 30 minutes, 1 hour and 2 hours, It currently gathers data from poloniex and computes features such as the orderbook imbalance, slopes in the weighted trade averages and spread’s among other indicators to do projection, It uses an evidence based algorithm that searches over 8 million historical examples of past market data every few seconds to find supporting evidence for its projections.

Lee says “Thats our secret sauce, it enables us to keep our algorithm responding to market conditions by feeding in new historical market data on the fly and keep it robust by feeding it varied market data, for example data from extreme swings like hardforks and coin launches”.

BitBank.nz has a referral/affiliate marketing program so anyone can signup and share their unique link to make 0.003 BTC for each paying user they refer.

Signup to BitBank.nz ’s all inclusive Cryptocurrency Forecasts plan starting at .03 BTC for 3 months.

Leave a Reply

Your email address will not be published. Required fields are marked *